search

victoresque / pytorch-template

PyTorch deep learning projects made easy.
MIT License
4.7k stars 1.08k forks source link

ZeroDivisionError in TensorboardWriter.set_step() #79

Open Andry-Bal opened 3 years ago

Andry-Bal commented 3 years ago

There is a risk of getting ZeroDivisionError: float division by zero in TensorboardWriter.set_step() method in line: self.add_scalar('steps_per_sec', 1 / duration.total_seconds()) I get this when running example config with "tensorboard": false

chiragthakur93 commented 3 years ago

I get the same Zerodivisionerror, even with "tensorboard": true Following is the error:

  File "G:\NewTemplate\pytorch-template-master\train.py", line 73, in <module>
    main(config)

  File "G:\NewTemplate\pytorch-template-master\train.py", line 54, in main
    trainer.train()

  File "G:\NewTemplate\pytorch-template-master\base\base_trainer.py", line 61, in train
    result = self._train_epoch(epoch)

  File "G:\NewTemplate\pytorch-template-master\trainer\trainer.py", line 68, in _train_epoch
    val_log = self._valid_epoch(epoch)

  File "G:\NewTemplate\pytorch-template-master\trainer\trainer.py", line 91, in _valid_epoch
    self.writer.set_step((epoch - 1) * len(self.valid_data_loader) + batch_idx, 'valid')

  File "G:\NewTemplate\pytorch-template-master\logger\visualization.py", line 48, in set_step
    self.add_scalar('steps_per_sec', 1 / duration.total_seconds())

ZeroDivisionError: float division by zero
sixian-C commented 1 year ago

I have the same issue, and I dont how to debug it.